On Mon, Sep 15, 2025 at 12:05 PM Eric Botcazou <botca...@adacore.com> wrote: > > > Yes please. Can we assert the MEM_REF offset is zero and the MEM_REF > > isn't type-punning, aka TREE_TYPE of the MEM_REF is compatible with > > the decls type, or is this not easily possible (again because of the > > placeholders)? > > Yes for the offset, but no for the type punning because we allow it for these > types (they have the "may_alias" attribute). > > > But the subprogram invocation would be guarded with a check for writability. > > How do you write a run-time check for writability in GENERIC though?
Well, if (writable_p) where the program appropriately sets that flag? > If the > program invokes a subprogram that contains an executable store on an object > that is a read-only memory, then execution is erroneous. Note that the doc of > TREE_THIS_NOTRAP suggests that it is orthogonal to readonly considerations. Ah, I wasn't aware of that. This makes TREE_THIS_NOTRAP possibly not usable for tree_could_trap_p :/ One could read the docs so that it means when you have a read with TREE_THIS_NOTRAP then you can't infer from that that writing to it is OK though. Rather than meaning that a write with TREE_THIS_NOTRAP meaning that write could still trap when the store is to readonly memory? > > The issue is that TREE_THIS_NOTRAP is "global" and not flow-sensitive. > > Yes, that's the reason for all this game. Understood. Note I'm completely unsure as to what extent we see TREE_THIS_NOTRAP being used (from frontends). I know OpenMP lowering uses it quite extensively, IIRC also nested function lowering. Grepping finds a single use outside of Ada, in go-gcc.cc Richard. > -- > Eric Botcazou > >
